New question about a different (this one, merely hypothetical at this point in time) animal behaviour study. It would involve running a negatively phototactic and positively thigmotactic species of invertebrate through a maze in an escape trial (when the animal finds the "hiding spot" in the otherwise bright and open maze the animal is rewarded by being allowed to sit in there undisturbed for a while). The idea is to see if there is any evidence of learning the route to the "hiding spot" as the trial is subsequently repeated. However, the animal species in question is fairly expensive. What is the minimum sample size you would suggest for this study? Statistical power calculations generally require things that are unknown to me at this point in time, such as standard deviation of the data, and extrapolating from trials of mice run through mazes and rewarded with food seems potentially problematic to me, so I thought I'd ask researchers that have experience in running these type of trials. Any help would be greatly appreciated.
Dear Nick,
Already when you have groups of 10 mice and have them learn a maze then one may be a slow learner and thus increase the SD which means that the difference between experimental and control mice does not necessary turn out significant. If you have two groups the min sample size is 6 animals (sign otr T-test), if you have four groups the min is 4 per group if you compute Friedmann's analysis of variance for dependent groups.
- Badges
- Science topic
- Similar topics
- Biological Science
- Zoology
- Invertebrate Zoology